amuseUSB is an intelligent, automated USB music loader that scans your existing songs, gathers metadata, generates AI-based personalized music recommendations, and batch-downloads new MP3 tracks directly into your USB drive with high speed and fully visible per-song progress bars.

Built with Python and designed for portability, amuseUSB turns any USB stick into a dynamically curated music library.

---

🚀 Features

• Reads existing USB music folders and extracts song metadata
• AI-powered recommendation generation (Gemini API expected)
• Parallel downloading (5 at a time) for maximum speed
• Per-song live progress bars using Rich
• FFmpeg-backed MP3 conversion (high-quality audio)
• Automatic retry system for unstable network links
• Clean and safe output file names
• Automatic USB directory handling
• Full logging and error reporting

---

📂 Project Structure

amuseUSB/
 ├── music_recommendations.json
 ├── phase1_scan_usb.py
 ├── phase2_get_recommendations.py
 ├── phase3_download_recommendations.py
 ├── README.md

---

🛠 Requirements

• Python 3.10+

• FFmpeg (must set --ffmpeg-location)

• Packages:

  yt_dlp
  rich
  

  Install:

  pip install yt-dlp rich

---

📥 Setup

1. Install FFmpeg
   Download from the official site and extract it somewhere permanent.

2. Locate FFmpeg path
   Example:

   D:\ffmpeg-8.0-essentials_build\bin
   

3. Update the script
   Set:

   FFMPEG_PATH = r"D:/ffmpeg-8.0-essentials_build/bin"
   USB_PATH = "E:/AI_Recommendations"

4. Place your music_recommendations.json
   Format:

   {
     "recommendations": [
       { "song": "Shape of You", "artist": "Ed Sheeran" },
       { "song": "Starboy", "artist": "The Weeknd" }
     ]
   }

You're ready to go.

---

▶ Running amuseUSB

python phase1_scan_usb.py #Step one
python phase2_get_recommendations.py #Step two
python phase3_download_recommendations.py #Step three

You will see progress bars like:

Searching: "Song Name"
Downloading: ███████░░░ 74%

Downloads are saved directly into your USB.

---

⚡ Performance

• Parallel downloads: 5
• Total speed: ~50--80 Mbps depending on network
• Average size per 100 HQ MP3 files: 0.7--1.1 GB

---

🔧 Configuration

Inside the script:

MAX_THREADS = 5
RETRY_LIMIT = 3
USB_PATH = "E:/AI_Recommendations"
FFMPEG_PATH = "D:/ffmpeg-8.0-essentials_build/bin"

Change these as needed.

---

🧑‍💻 Author

Dewashish Lambore

---

📄 License

This project is open for personal use and modification. Attribution appreciated.

---

⭐ Support

If you enjoyed this project, consider starring the GitHub repo!